The relationship between adipokine levels and bone mass—A systematic review

Abstract Introduction Adipose tissue is the source of a broad array of signalling molecules (adipokines), which mediate interorgan communication and regulate metabolic homeostasis. Alterations in adipokine levels have been causally implicated in various metabolic disorders, including changes in bone mass. Osteoporosis is the commonest progressive metabolic bone disease, characterized by elevated risk of fragility fractures as a result of a reduced bone mass and microarchitectural deterioration. The effects of different adipokines on bone mass have been studied in an attempt to identify novel modulators of bone mass or diagnostic biomarkers of osteoporosis. Methods In this review, we sought to aggregate and assess evidence from independent studies that quantify specific adipokines and their effect on bone mineral density (BMD). Results A literature search identified 57 articles that explored associations between different adipokines and BMD. Adiponectin and leptin were the most frequently studied adipokines, with most studies demonstrating that adiponectin levels are associated with decreased BMD at the lumbar spine and femoral neck. Conversely, leptin levels are associated with increased BMD at these sites. However, extensive heterogeneity with regards to sample size, characteristics of study subjects, ethnicity, as well as direction and magnitude of effect at specific skeletal anatomical sites was identified. The broad degree of conflicting findings reported in this study can be attributed several factors. These include differences in study design and ascertainment criteria, the analytic challenges of quantifying specific adipokines and their isoforms, pre‐analytical variables (in particular patient preparation) and confounding effects of co‐existing disease. Conclusions This review highlights the biological relevance of adipokines in bone metabolism and reinforces the need for longitudinal research to elucidate the causal relationship of adipokines on bone mass.


| INTRODUC TI ON
Adipose tissue is an active endocrine organ and the source of several signalling molecules known as adipokines. It is comprised of a heterogeneous cell population consisting of stromal cells, fibroblasts, macrophages, preadipocytes and adipocytes. 1 Adipokines have been implicated in the regulation of numerous metabolic processes, including energy metabolism, satiety and interorgan communication. 2 Dysregulated levels of several adipokines are observed in obesity and type 2 diabetes mellitus (T2DM) and are associated with cardiometabolic traits. This is likely due to the varying effect of different adipokines on chronic inflammation. For example, leptin stimulates the production of various proinflammatory cytokines, whereas adiponectin exerts an inhibitory effect on TNFα-induced activation of nuclear factor kappa B (NF-kB). 3 In addition, adipokines have been variably implicated in osteoporosis. This is a complex progressive metabolic bone disease characterized by microarchitectural deterioration of bone tissue and a reduction in bone mass and strength leading to an increased fracture risk. 4 In obesity, an increased bone turnover accompanied by a high bone mineral density (BMD) is typically observed. This is an adaptive response to an enlarged body frame potentially modulated by changes in body weight, mechanical loading, differences in bone remodelling, as well as age and gender-specific changes in body composition. 5 The role of dysregulated adipokine levels in this process is not clearly understood, and establishing a causal direction remains challenging given the complex and multifactorial nature of the obesity-BMD interaction. 6 Receptors for adiponectin, one of the most abundant adipokines in the circulation, are expressed on bone lineage cells. 7 Obese individuals have reduced circulating levels of adiponectin. This adipokine limits hepatic gluconeogenesis and promotes muscle insulin sensitivity, and a reduction in circulating adiponectin concentrations is reported in obese individuals. 8 The link between adiponectin and bone physiology is less well understood. Mouse models indicate that adiponectin enhances osteoblastogenesis and promotes bone repair. 9 In healthy humans, this association holds true. However, the presence of menopause, obesity, metabolic syndrome and other chronic inflammatory states reverses adiponectin's protective effect on bone mass, favouring increased bone resorption. 8 Leptin mediates the complex crosstalk between the central nervous system (CNS), adipose tissue and energy homeostasis. Its concentration positively correlates with body fat. The expression of leptin receptors on chondrocytes and osteoblasts hints towards its role in bone turnover and endochondral ossification. 10 Lep −/− mice exhibit obesity and reduced femoral neck BMD, bone volume and cortical thickness. 11 Human studies similarly demonstrate its importance in the maintenance of bone structural integrity, with lower leptin levels linked to abnormal bone microarchitecture and increase fracture risk. 12 Leptin exhibits angiogenic properties and promotes bone formation via the downregulation of receptor activator of nuclear factor kappa-Β ligand (RANKL). 13,14 Additionally, leptin activates RANKL's decoy receptor osteoprotegerin (OPG) and fibroblast growth factor 23 (FGF23), further limiting bone resorption. 15,16 Gender alters the association between leptin and BMD, with some studies finding non site-specific correlations between leptin and BMD in females (regardless of menopausal status) but not in males, which can be attributed to the differences in leptin production and leptin sensitivity in males and females. 17 Resistin is another adipokine that regulates adipogenesis and multiple inflammatory and metabolic processes, although its mechanism of action is not yet fully understood. 18 Resistin increases the expression of proinflammatory cytokines such as IL-1, IL-6 and TNFα, leading to an increased production of reactive oxygen species.
Although these effects link resistin to cardiovascular dysfunction, 19 few studies have been carried out with regards to the association of resistin with bone health, likely due to the relative novelty of this adipokine.
In summary, adipokines are important modulators of bone mass and may demonstrate potential use as bone-related disease biomarkers. This review aggregates and evaluates evidence from individual studies investigating the effect of adipokines on BMD.  Table S1; (B) Adiponectin levels in individuals with comorbidities affecting bone mass based on studies described in Table S2. The first row indicates the study from which the data were derived. The column to the right denotes the site/s of BMD measurement. The studies highlighted in green were longitudinal studies, whilst the rest were case-control studies. Cells highlighted in blue indicate a positive relationship between BMD and adiponectin. Cells highlighted in red indicate a negative relationship between BMD and adiponectin. Cells highlighted in grey indicate no significant findings. Cells left blank were not tested. BUA, broadband ultrasound attenuation; FN, femoral neck; LS, lumbar spine; TB, total body; TH, total hip; TS, thoracic spine. Table S1; (B) Leptin levels in individuals with comorbidities affecting bone mass based on studies described in Table S2. The first row indicates the study from which the data were derived. The column to the right denotes the site/s of BMD measurement. The studies highlighted in green were longitudinal studies, whilst the rest were case-control studies. Cells highlighted in blue indicate a positive relationship between BMD and leptin. Cells highlighted in red indicate a negative relationship between BMD and leptin. Cells highlighted in grey indicate no significant findings. Cells left blank were not tested. BUA, broadband ultrasound attenuation; FN, femoral neck; LS, lumbar spine; TB, total body; TH, total hip. heterogeneity in the direction and magnitude of the relationship between the adipokines and BMD.

F I G U R E 2 Association of serum leptin levels with BMD at different anatomical sites. (A) Leptin levels in individuals with no comorbidities based on studies described in
Numerous confounding variables may affect the observations presented by the studies outlined in this review, particularly the study designs implemented, leading to discordant results. Meng et al. 24 assessed the effects of leptin levels on BMD using a Mendelian randomization approach, with genetic variants that lead to increased leptin levels resulting in a reduced LS BMD. The same study did not identify consistent associations at different anatomical sites, which may be a contributing factor to the heterogeneity between studies.  30 Additionally, genetic variation in ADIPOQ and LEP genes has been shown to alter adipokine production and function. 31 Importantly, the risk and progression of osteoporosis as well as the chronic subclinical inflammatory responses characteristic of metabolic diseases are strongly modulated by complex gene-environment interactions that are challenging to quantify. Various drugs such as metformin and statins also modulate inflammatory responses. 32,33 The direct assessment of adipokine levels in biologically relevant tissue, such as bone matrix, is a potentially superior alternative method to their determination in blood, which overcomes the limitations imposed by improper patient preparation and differences in circulating adipokine half-life concentrations. Notwithstanding, bone matrix is less accessible to sampling and not ideal for biomarker studies, particularly as it requires extensive processing due to the biological complexity of the tissue. 34 Some of the study populations investigated in this work were also diagnosed with different comorbidities or underwent interventions (Table S2)  This can provide a potential cause for the incongruent results observed in the studies described above. 39 Whilst no significant changes in adiponectin concentration can be attributed to the fasting status of the participant, 40 studies have shown that the same cannot be said for leptin, being significantly reduced in fasted individuals (intermittent and alternate day fasting) as well as those on energy-restricted diets. 41 Most of the reviewed studies are based on the analysis of fasting individuals; nonetheless others included non-fasting individuals which may have inadvertently contributed to the discrepant observations. Other preanalytical variables that may significantly alter the measurement of adipokine levels include the effects of medication on the assessed adipokines, as well as the effects of long-term storage and freezing. In some of the reviewed studies, the study population was too small to be able to reach the necessary statistical power to ascertain the observations noted. This is especially evident in studies listed in Table S2. Weaker associations between adipokines and BMD may remain unnoticed through the use of smaller populations. 42 This review shows that a causal direction between adipokines and BMD cannot be robustly inferred. Large biobank-driven longitudinal cohort studies could facilitate a better understanding of the role of these adipokines with regards to bone health.

CO N FLI C T O F I NTE R E S T
All authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available in the public domain MEDLINE searchable via PubMed at https://pubmed. ncbi.nlm.nih.gov/.

E TH I C A L A PPROVA L
Ethics approval was not required since this is a systematic review of previously published studies.